Bluelibra Use PSl-BLAST (http://www.ncbi.nlm.nih.gov/BLAST/Blast.cgi) to search
ID: 1894004 • Letter: B
Question
Bluelibra
Use PSl-BLAST (http://www.ncbi.nlm.nih.gov/BLAST/Blast.cgi) to search for sequences related to the C chain of the 1KCG crystal structure (you can find its sequence on the RCSB PDB). What protein is this To what class of proteins is homology revealed with PSl-BLAST? How do the identity and ordering of results change after a second and third iteration of the PSl-BLAST algorithm? (Do this by clicking on the 'Go' button.) With what you know about PSI-BLAST from the lectures, why do you think this change occurs?Explanation / Answer
At http://www.ncbi.nlm.nih.gov/BLAST/Blast.cgi, launch Protein Blast
As the accession number, enter the pdb entry 1KCG_C
The letter C means chain ID labeled C.
Further down the page, select PSI-BLAST, then execute BLAST
Wait at least 30 seconds. It will display 6 or seven windows, slowly, then settle with the display of the protein sequence alignment.
At the PSI BLAST Iteration 1, it identified this entry as Chain C, NKG2D in complex with ULBP3. Chain C denotes the 178 residues ULBP3 part of it.
b) In the Description table, there are many entries listed, some of which refering to this chain C belonging to the MHC class I superfamily. Note multiple entry listed are for either hypothetical or predicted sequences. (The symbol M next to the first entry points to the human chromosome entry 6q25.) At the PSI-BLAST results, further down and below the descriptions table, the partial sequence alignments decreases in %homology as one looks down the list. Running PSI-BLAST again will help focus on the consistency of the entries.
c) Run PSI-BLAST Iteration 2 with max 500 cycles. The GO button is located in the Descriptions box. Wait ......... even if it seems to do nothing....
The results of that second round shows some consistency. Looking further down, the sequence alignemnt is not as high from the fourth entry down, but the results appear more consistent with each other in their descriptions. The results and statistics identify other homologous entries like UL16 binding protein 2 [Pan troglodytes], retinoic acid early transcript 1L [Homo sapiens], UL16 binding protein 2, isoform CRA_b [Homo sapiens], etc.
The second round of PSI-BLAST excludes some of the identical sequences and thus help identify more distant relatives like isoforms, homologous sequences, alternative transcripts, etc. This alignment is better tailored to the original chain C. There is just so much that can be displayed with only 100 Blast Hits..
At the next PSI-BLAST iteration, the hits descriptions are now quite consistent, identifying this family of sequences as MHC Class I antigen. Note the lower %homology of all these sequences (see further down the page). However, the sequences increased in their consistency. That's what PSI-BLAST does.